Line ending
ⓘ
What every line break in pasted text becomes on the wire. Clipboard text on Linux is LF-only, which the MicroBeast does not take as a line break at all — hence the CR default. A CRLF pair in the clipboard counts as one break, never two. This is a separate setting from Settings ▸ Enter key sends: that one governs the Enter key, this one governs paste, and the two can hold different values. "As-is" passes the clipboard bytes through untouched.
CR (0x0D)
LF (0x0A)
CRLF (0x0D 0x0A)
As-is (no change)
Chunk size
ⓘ
How many bytes go out back-to-back before the pump waits. Every chunk is exactly this size except the last, whatever the bytes contain. Burst size is not what governs delivery: on real hardware 1 byte every 100 ms and 2 bytes every 200 ms are the same 10 B/s and behave identically, so it is the throughput below that matters. Bigger chunks at the same throughput simply mean fewer, larger bursts.
1 byte
2 bytes
4 bytes
8 bytes
16 bytes
32 bytes
Pause
ⓘ
Idle time between chunks, giving the receiver room to drain. Measured on a real MicroBeast with no flow control: 1 byte every 200 ms — 5 B/s — delivers an 800 B block into VIBE intact, and 10 B/s only nearly works. 150 ms is about 6.7 B/s, in the gap between them. "None" means no pause at all: the wire is the only limit.
None (wire speed)
5 ms
10 ms
20 ms
50 ms
100 ms
150 ms
200 ms
Throughput
ⓘ
What the two rows above add up to — chunk size divided by pause. It is shown, never set: you describe the cadence and read the rate off it. When the open port was opened with hardware flow control the pump does not pace at all, because the receiver handshakes per byte and that beats any fixed cadence — the two rows keep their values, and they apply again the moment a bare port is opened, but they are not in force and this line says so.
≈ 5 B/s
Pacing applies unless the open port uses hardware flow control — including when nothing is connected. The paste chip reports the rate actually achieved while a paste runs.